None identical results from differents run

[remilamoureuxlevesque]

I am running a benchmark for evaluating the best algorithm to use for my application, but I cannot get twice the same result from running the benchmark two times in a row.

Does it have somethings in the algorithm that could be related to that ? or are you able to get repetitive results from algorithms in pyOD ?

Thanks for you feedback.

[yzhao062]

Many od algorithms have built in randomness

[remilamoureuxlevesque]

Oh ok,

Is there a way to remove or reduce the effect of the randomness ?

[yzhao062]

it depends on algorithms, knn lof is deterministic. isolation forest is not

[remilamoureuxlevesque]

Ok thank you for your anwser, I will look into that.